Drift control method and unmanned vehicle

Abstract

This disclosure provides a drift control method and an unmanned vehicle, relating to the field of autonomous driving technology. The method includes: predicting the desired balance parameters and feedforward control parameters of the vehicle at the target location based on road information and vehicle state information; generating feedback control parameters with the goal of controlling the vehicle's current balance parameters to the desired balance parameters; and controlling the vehicle to drift at the target location based on the feedforward and feedback control parameters. This disclosure can control a vehicle to drift on curves, significantly improving the cornering efficiency of trucks. Simultaneously, the control strategy based on the feedforward and feedback control parameters can also enable the vehicle to precisely maintain its drift posture during drifting, avoiding the risk of instability during the drift process.

Description

Technical Field

[0001] This disclosure relates to the field of autonomous driving technology, and in particular to a drift control method and an unmanned vehicle. Background Technology

[0002] In unmanned mining scenarios, vehicles need to operate efficiently in complex road conditions and extreme environments. Especially when driving on curves and slopes in mines, the vehicle's cornering speed and stability are key factors affecting transportation efficiency and safety.

[0003] In vehicle cornering control, related technologies typically rely on a conservative, speed-reducing cornering strategy. This approach ensures vehicle stability by limiting steering angle and driving force. However, this method often proves too conservative in complex road conditions, resulting in insufficient utilization of the vehicle's dynamic boundaries and limited cornering speed. Summary of the Invention

[0004] To overcome the problems existing in related technologies, this disclosure provides a drift control method and an unmanned vehicle.

[0005] According to a first aspect of the present disclosure, a drift control method is provided, comprising: Based on road information and vehicle status information at the target location, predict the desired balance parameters and feedforward control parameters of the vehicle at the target location; With the goal of controlling the current balance parameters of the vehicle to the desired balance parameters, feedback control parameters are generated; Based on the feedforward control parameters and the feedback control parameters, the vehicle is controlled to drift at the target position.

[0006] In some embodiments, before generating feedback control parameters with the objective of controlling the vehicle's current balance parameters to the desired balance parameters, the method further includes: If the vehicle meets the drift mode triggering conditions, it is determined that the vehicle has entered drift mode, and the step of generating feedback control parameters with the goal of controlling the current balance parameters of the vehicle to the desired balance parameters continues to be executed; The drift mode triggering condition includes at least one of the following: The predicted wheel sideslip parameter of the vehicle at the target position is greater than the wheel sideslip parameter threshold, wherein the wheel sideslip parameter threshold is determined based on the slippery data of the target position, and the predicted wheel sideslip parameter is determined based on the current balance parameter of the vehicle combined with the road information of the target position. The wheel sideslip parameter is used to characterize the trend of the vehicle's steering attitude change. The deviation between the vehicle's current attitude information and the target attitude information satisfies a preset deviation condition.

[0007] In some embodiments, the slippery data of the target location is obtained in the following manner: Based on historical slippery data of the vehicle's driving area, preliminary slippery information of the driving area is generated; Based on the current environmental information, the preliminary slippery information is corrected to generate the current slippery information of the driving area; wherein, the environmental information includes at least one of weather information and road surface condition information; Determine the slippery data that matches the target location from the current slippery information.

[0008] In some embodiments, the vehicle's current attitude information includes: current yaw rate; the target attitude information includes: target yaw rate; and the preset deviation condition includes: the current yaw rate and the target yaw rate are in the same direction, and the current yaw rate is greater than a yaw rate threshold; and / or The vehicle's current attitude information includes: current lateral position information and current heading information. The target attitude information includes: target lateral position information and target heading information. The preset deviation conditions include: the lateral position error and the heading error are in opposite directions, and the heading error is greater than the heading error threshold. The lateral position error is determined based on the current lateral position information and the target lateral position information, and the heading error is determined based on the current heading information and the target heading information.

[0009] In some embodiments, controlling the vehicle to drift at the target position based on the feedforward control parameters and the feedback control parameters includes: The control quantity shown by the feedforward control parameters is superimposed with the control quantity shown by the feedback control parameters to generate the target control quantity; The target control quantity is converted into a control signal to control the vehicle to drift at the target position.

[0010] In some embodiments, generating feedback control parameters with the goal of controlling the current balance parameters of the vehicle to the desired balance parameters includes: Construct a target state quantity sequence, wherein the target state quantity sequence is used to show the tracking target of the vehicle state parameters at each time in the prediction time domain, the vehicle state parameters include vehicle balance parameters, and the tracking target of the vehicle balance parameters is related to the expected balance parameters; The vehicle state parameters at the current moment and the target state sequence are input into a pre-built model prediction controller, so that the model prediction controller generates the feedback control parameters at the current moment of the vehicle with the objective of minimizing the difference between the vehicle state parameters at each moment of the vehicle and the corresponding tracking target.

[0011] In some embodiments, constructing a target state quantity sequence includes: Determine the difference between the current balance parameters of the vehicle and the desired balance parameters; When the difference meets the preset conditions, starting from the current balance parameter of the vehicle, the tracking target of the vehicle balance parameter at each moment in the prediction time domain is generated by sequentially increasing the preset transition rate, so as to smoothly transition the tracking target of the vehicle balance parameter from the current balance parameter to the desired balance parameter; wherein, the preset transition rate is positively correlated with the rate of change of road curvature of the vehicle's driving position in the prediction time domain. Based on the vehicle balance parameters at each time point, a target state quantity sequence is constructed.

[0012] In some embodiments, the method further includes: If the vehicle no longer meets the drift mode triggering conditions or the vehicle exits the curve, the vehicle is determined to have ended drifting, and the vehicle's current balance parameters are smoothly transitioned to the preset balance parameters in normal driving state, so that the vehicle smoothly transitions from drifting to normal driving.

[0013] In some embodiments, predicting the desired balance parameters and feedforward control parameters of the vehicle at the target location based on road information and vehicle state information includes: A system state model of the vehicle is constructed, which is associated with the vehicle's longitudinal velocity, center of gravity sideslip angle, yaw rate, and feedforward control parameters. The road information and the state information are input into the system state model. Under the condition that the rate of change of longitudinal velocity, the rate of change of centroid sideslip angle and the rate of change of yaw rate in the system state model are constrained to zero, the desired balance parameters and feedforward control parameters of the vehicle at the target position are determined by the system state model. The desired balance parameters include the centroid sideslip angle.

[0014] According to a second aspect of this disclosure, an unmanned vehicle is provided for performing the drift control method described in the first aspect.

[0015] According to a third aspect of this disclosure, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a controller, implements the method described in the first aspect.

[0016] According to a fourth aspect of this disclosure, a computer program product is provided, comprising a computer program that, when executed by a controller, implements the steps of the method described in the first aspect.

[0017] The solution provided in this disclosure can predict the desired balance parameters and feedforward control parameters of the vehicle at the target location based on road information and vehicle state information. Then, with the goal of controlling the vehicle's current balance parameters to the desired balance parameters, feedback control parameters are generated. Based on the feedforward and feedback control parameters, the vehicle can be controlled to drift at the target location. This disclosure can control the vehicle to drift in curves, significantly improving the cornering efficiency of trucks. Simultaneously, the control strategy based on the feedforward and feedback control parameters can also enable the vehicle to precisely maintain its drift posture during drifting, avoiding the risk of instability. Attached Figure Description

[0018] Figure 1 A flowchart illustrating a drift control method according to an embodiment of this disclosure is shown.

[0019] Figure 2 The diagram illustrates a flowchart of a method for predicting desired balance parameters and feedforward control parameters according to an embodiment of this disclosure.

[0020] Figure 3 The diagram shows a flowchart of a method for determining feedback control parameters according to an embodiment of this disclosure.

[0021] Figure 4 The diagram shows a flowchart of a method for constructing a target state variable sequence according to an embodiment of the present disclosure.

[0022] Figure 5 A schematic diagram of a drift control device according to an embodiment of the present disclosure is shown. Detailed Implementation

[0023] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0024] First, this disclosure provides a drift control method that can be executed by any unmanned vehicle.

[0025] Figure 1 This diagram illustrates a flow chart of a drift control method according to an embodiment of the present disclosure, as shown below. Figure 1 As shown, the method includes the following steps S101 to S103.

[0026] S101, based on road information and vehicle state information at the target location, predicts the vehicle's desired balance parameters and feedforward control parameters at the target location.

[0027] In some embodiments, the target location can be a pre-aiming point on the vehicle's trajectory, which can be set based on time or distance. For example, a path point at a preset distance from the vehicle in its current direction of travel can be selected as the pre-aiming point, or a path point the vehicle will reach after a preset time can be selected as the pre-aiming point. In this embodiment, the pre-aiming point is used to plan the drift posture in advance, avoiding lag in vehicle posture response.

[0028] Road information includes the road curvature at the target location (pre-aiming point), and can also be combined with the road surface adhesion coefficient to provide working condition basis for solving the equilibrium parameters and control parameters.

[0029] Vehicle status information can cover core dynamic parameters such as the vehicle's current longitudinal speed, and can be collected in real time through onboard inertial navigation systems, wheel speed sensors, and other equipment to reflect the vehicle's current attitude.

[0030] In some embodiments, a zero-dynamic equation can be constructed by combining the vehicle dynamics model with the core concept of maintaining steady-state drift when the vehicle reaches the target position. The zero-dynamic equation ensures that the solution represents the steady-state drift equilibrium point at the target position, constrained by the zero rate of change of some vehicle dynamic parameters during drift. Specifically, this equation takes the desired longitudinal velocity and desired yaw rate at the equilibrium point as inputs. By solving this equation, the desired equilibrium parameters and feedforward control parameters of the vehicle during steady-state drift at the target position can be predicted. The desired longitudinal velocity can be the vehicle's current longitudinal velocity, and the desired yaw rate can be calculated by combining the desired longitudinal velocity and the road curvature at the target position.

[0031] It should be noted that the core of the predicted desired balance parameter can be the desired sideslip angle, which enables the vehicle to drift steadily at the target position. This parameter is closely related to the stability of the vehicle's drift posture. The feedforward control parameter can be understood as the control parameter that works in conjunction with the desired balance parameter. It can include the desired front wheel steering angle and the desired rear wheel driving force. By controlling the front wheel angle when the vehicle reaches the target position to the desired front wheel steering angle and the rear wheel driving force to the desired rear wheel driving force, the vehicle's sideslip angle can be indirectly controlled to the desired sideslip angle, thereby maintaining a stable drift posture.

[0032] S102 generates feedback control parameters with the goal of controlling the vehicle's current balance parameters to the desired balance parameters.

[0033] In some embodiments, real-time deviations between the vehicle's current balance parameters and desired balance parameters can be eliminated by generating feedback control parameters, ensuring that the vehicle accurately approaches the drift posture of the target position to maintain drift stability.

[0034] For example, feedback control parameters can be generated using a Model Predictive Control (MPC). Specifically, the MPC can be constructed using the deviation between the vehicle's current balance parameters and the desired balance parameters as input. By predicting attitude changes in the prediction time domain, the feedback control parameters can be optimized and solved. The core function of the feedback control parameters is to dynamically correct deviations caused by real-time operating condition disturbances (such as fluctuations in road adhesion coefficient or slight trajectory deviations), ensuring that even with operating condition fluctuations, the vehicle can still stably track the desired balance parameters and avoid drifting and instability.

[0035] For example, the generated feedback control parameters correspond to the feedforward control parameters and may include the compensation amount for the desired front wheel steering angle and the compensation amount for the desired rear wheel driving force.

[0036] S103 controls the vehicle to drift at the target position based on feedforward control parameters and feedback control parameters.

[0037] In some embodiments, feedforward control parameters can be used as the control reference, and feedback control parameters can be used as dynamic compensation parameters. The final control signal is output after fusing the feedforward and feedback control parameters. The feedforward control parameters ensure that the overall vehicle attitude conforms to the steady-state drift requirement of the target position, avoiding large jumps in the control quantity. The feedback control parameters make small adjustments to address real-time deviations, compensating for errors caused by model approximation and operating condition disturbances.

[0038] For example, the control quantity shown by the feedforward control parameters can be superimposed with the control quantity shown by the feedback control parameters to generate a target control quantity. The target control quantity is then converted into a control signal to control the vehicle to drift at the target position.

[0039] The final fused control signal is sent to the vehicle's actuators to control the vehicle's attitude to gradually approach the attitude indicated by the desired equilibrium parameters, ensuring that the vehicle maintains a stable drift state when it reaches the target position. Furthermore, each control cycle of the vehicle can repeat the above steps, achieving closed-loop iterative control of the vehicle during the drift process, continuously adapting to changes in the road environment and attitude deviations during the drift.

[0040] This disclosed embodiment can pre-set a baseline for the drift attitude by predicting the desired balance parameters and feedforward control parameters of the target position (pre-aiming point). This can effectively compensate for the attitude response lag caused by heavy load and large inertia of the vehicle, and avoid attitude oscillations caused by untimely control. Furthermore, by integrating the control mechanism of feedforward control parameters and feedback control parameters, the tracking error of the vehicle balance parameters can be controlled within a very small range, effectively reducing the risk of sideslip and rollover during the drift process.

[0041] Therefore, this disclosure enables stable drift control of vehicles, allowing them to safely navigate corners without significantly reducing their speed, thus greatly improving the efficiency of vehicles when cornering.

[0042] The above embodiments have provided a detailed explanation of the technical concept of vehicle drift control disclosed in this disclosure. For ease of understanding, the following will be combined with... Figure 2 This disclosure details the prediction methods for the desired balance parameters and feedforward control parameters.

[0043] Figure 2 A flowchart illustrating a method for predicting desired balance parameters and feedforward control parameters according to an embodiment of this disclosure is shown. Figure 2 As shown, the method includes the following steps.

[0044] S201, Construct the system state model of the vehicle.

[0045] The system state model is associated with the vehicle's longitudinal velocity, center of gravity sideslip angle, yaw rate, and feedforward control parameters.

[0046] S202, road information and state information are input into the system state model, and the desired balance parameters and feedforward control parameters of the vehicle at the target position are determined by constraining the rate of change of longitudinal velocity, the rate of change of centroid sideslip angle and the rate of change of yaw rate in the system state model to zero.

[0047] The desired balance parameters include the centroid sideslip angle.

[0048] In some embodiments, by analyzing the force and torque balance of the vehicle during the drifting process, vehicle dynamics equations in the longitudinal, lateral, and yaw directions can be established to describe the changes in the vehicle's state over time.

[0049] Furthermore, since the core objective of drift control is to allow the vehicle to enter and maintain a steady-state drift, in a steady state, the vehicle's longitudinal velocity, sideslip angle, and yaw rate no longer change with time. Therefore, based on this characteristic, the vehicle dynamics equations can be expressed as descriptions of the rates of change of longitudinal velocity, sideslip angle, and yaw rate, thus obtaining the system state model.

[0050] For example, the system state model can be represented by the following equation: ; ; ; in, The longitudinal velocity change rate, The rate of change of the centroid sideslip angle. The rate of change of yaw angular velocity, For longitudinal velocity, The sideslip angle is the angle of the centroid. The yaw rate is angular velocity. This refers to the front wheel steering angle. Rear-wheel drive The vehicle dynamics equations used to represent the rate of change of longitudinal velocity are as follows: The vehicle dynamics equations used to represent the rate of change of the sideslip angle at the center of gravity are as follows: This is the vehicle dynamics equation used to represent the rate of change of yaw rate.

[0051] Specifically, after substituting the vehicle dynamics equations, the above system state model can be expressed as follows: ; in, For vehicle quality, For rotational inertia, The distance from the center of gravity to the front axle of the vehicle. This is the distance from the center of gravity to the rear axle of the vehicle. The lateral force is from the front wheel. This is the lateral force on the rear wheel.

[0052] By constraining the rates of change of longitudinal velocity, the rate of change of sideslip angle, and the rate of change of yaw rate in the system state model to zero, i.e. This allows us to obtain the zero dynamic equation, which in turn enables us to solve for the vehicle's desired balance parameters and feedforward control parameters at the target position.

[0053] In the solution process, the above formula... The desired longitudinal velocity at the target position equilibrium point can be obtained from the vehicle's state information. During the solution process, the longitudinal velocity currently collected by the sensor can be substituted into the solution. It can be obtained by combining the desired longitudinal velocity with the road curvature in the road information at the target location, that is, the desired yaw rate at the equilibrium point of the target location. and It can be calculated from the tire slip angle using a tire model (e.g., a brush tire model).

[0054] By solving the zero dynamic equation, the desired balance parameters of the vehicle at the target position can be determined (i.e., in the above equation). ) and feedforward control parameters (i.e., in the above formula) and The solution method for the zero dynamic equation can be Newton's iteration method, Gauss-Newton method, quasi-Newton method, alternating multiplier method, etc., which will not be elaborated in the embodiments of this disclosure.

[0055] This disclosed embodiment proactively solves the steady-state drift equilibrium point using zero dynamic equations, enabling vehicles to drift actively under extreme conditions such as low-adhesion curves. This allows the vehicle to maintain higher speeds in curves or shorten single-trip transport time through optimized routes (such as reducing turning radius). In unmanned mining truck applications, this directly increases the amount of ore transported per unit time, precisely meeting the core demands of cost reduction and efficiency improvement in mine operations, resulting in direct and considerable economic benefits.

[0056] The above embodiments have provided a detailed explanation of how to determine the feedforward control parameters. It can be understood that the final control signal used to control vehicle drift is obtained by fusing the feedforward control parameters and the feedback control parameters. Therefore, after introducing the method for determining the feedforward control parameters, the following will combine... Figure 3 This further explains how the feedback control parameters are determined.

[0057] Figure 3 This diagram illustrates a flowchart of a feedback control parameter determination method according to an embodiment of the present disclosure. Figure 3 As shown, the method includes the following steps.

[0058] S301, Construct the target state variable sequence.

[0059] The target state quantity sequence is used to show the tracking target of the vehicle state parameters at each time in the prediction time domain. The vehicle state parameters include the vehicle balance parameters, and the tracking target of the vehicle balance parameters is related to the expected balance parameters.

[0060] S302, the vehicle state parameters and target state sequence at the current moment are input into the pre-built model predictive controller, so that the model predictive controller generates the feedback control parameters at the current moment of the vehicle with the goal of minimizing the difference between the vehicle state parameters at each moment of the vehicle and the corresponding tracking target.

[0061] In some embodiments, the target state quantity sequence can be understood as the tracking target sequence of the model predictive controller, used to define the state target that the vehicle needs to approach at each time in the future prediction time domain.

[0062] For example, vehicle state parameters may include lateral displacement, heading, longitudinal velocity, sideslip angle, and yaw rate. The target state quantity sequence can be understood as the tracking target at various times in the future prediction time domain for the aforementioned vehicle parameters.

[0063] Among them, the centroid sideslip angle can be used as the vehicle balance parameter mentioned above. Accordingly, the tracking target at each moment of the centroid sideslip angle is the target quantity associated with the desired balance parameter (i.e., the desired centroid sideslip angle) mentioned above. As for other parameters besides the centroid sideslip angle, the tracking targets of these parameters at each moment can be obtained based on the current vehicle state information combined with the trajectory decision information upstream of the unmanned vehicle, which will not be elaborated in this embodiment.

[0064] During the process of controlling vehicle drift, the difference between the vehicle state parameters at each moment and the corresponding tracking target can be minimized to make the vehicle state parameters approximate the corresponding tracking target, thereby ensuring that the vehicle attitude continuously converges to a stable drift attitude.

[0065] In some embodiments, the actual vehicle state parameters at the current moment can be input together with the target state sequence into a pre-built model predictive controller. The model predictive controller can take minimizing the difference between the vehicle state parameters at each moment and the corresponding tracking target in the future prediction time domain as its core objective. Through online optimization, it obtains the optimal control sequence and outputs the control parameters (including the front wheel steering angle and the rear wheel driving force) corresponding to the current moment in the control sequence as feedback control parameters, thereby dynamically adjusting the deviation between the current vehicle state and the target state.

[0066] Specifically, the objective function of the model predictive controller can be expressed as: ; in, for The tracking target corresponding to the vehicle state parameters at any given time can be obtained through the target state variable sequence. for The actual vehicle state parameters corresponding to the given time. for The control parameters corresponding to each moment, To predict the time domain, To control the time domain, The state weight matrix is... To control the weight matrix, For time steps, The objective function is denoted as .

[0067] Vehicle state parameters in the objective function ,in, This is lateral displacement. For the course, For longitudinal velocity, The sideslip angle is the angle of the centroid. Yaw rate; control parameters ,in, This refers to the front wheel steering angle. It provides rear-wheel drive.

[0068] The state-space equations are: ; in, This is the system state error quantity. and For the model predictive controller at the current operating point The Jacobian matrix is ​​calculated at point .

[0069] For example, in order to ensure numerical stability under different solution step sizes, the yaw rate involved in the solution process is... and centroid side slip angle The differential part can be discretized using the backward Euler method.

[0070] Understandably, the model predictive controller generates feedback control values ​​at various time points through rolling optimization, which are used to compensate for errors in the feedforward control values. By fusing the corresponding parameters from the feedforward control parameters with the feedback control parameters at the current time point, the complete target control value for the vehicle at that current time point can be obtained.

[0071] Next, we will describe in detail the correlation between the tracking target and the desired centroid sideslip angle involved in the above steps of constructing the target state variable sequence. The core of this correlation is the pursuit of smooth transition under large deviations between the current balance parameter (current centroid sideslip angle) and the desired balance parameter (desired centroid sideslip angle), and the pursuit of control efficiency under small deviations.

[0072] Specifically, the difference between the current centroid sideslip angle and the desired centroid sideslip angle can be determined first. .in, The difference. This is the current centroid sideslip angle. The desired centroid sideslip angle.

[0073] like Then let the tracking target at each time step in the target state sequence be... ;like Then let the tracking target at each time step in the target state sequence be... With transition rate From the current centroid side slip angle Side slip angle towards the desired centroid Transition. Among them, the transition rate... With the rate of change of road curvature Positive correlation and For different centroid sideslip angle difference thresholds, .

[0074] In other words, In this case, the difference between the current centroid sideslip angle and the desired centroid sideslip angle is relatively small. At this point, it can be determined that the vehicle's current attitude is close to a stable drift attitude, so there is no need for a gradual transition, and the pursuit of control efficiency can be prioritized.

[0075] And in In this case, the difference between the current centroid sideslip angle and the desired centroid sideslip angle is relatively large. If the current centroid sideslip angle is directly made to track the desired centroid sideslip angle, it may cause a sudden change in the vehicle's attitude. Therefore, it is necessary to gradually make the current centroid sideslip angle approach the desired centroid sideslip angle through a gradual transition to avoid sudden changes in attitude that could cause the vehicle to become unstable.

[0076] Based on the above technical concept, the method for constructing the target state quantity sequence containing vehicle balance parameters (center of gravity sideslip angle) in the embodiments of this disclosure can be described as follows: Figure 4 The method shown.

[0077] Please refer to Figure 4 , Figure 4 This diagram illustrates a flowchart of a method for constructing a target state variable sequence according to an embodiment of this disclosure. Figure 4 As shown, the method includes the following steps: S401, determine the difference between the vehicle's current balance parameters and the desired balance parameters.

[0078] S402, if the difference meets the preset conditions, starting from the current balance parameter of the vehicle, the tracking target of the vehicle balance parameter at each moment in the prediction time domain is generated by increasing the preset transition rate sequentially, so as to smoothly transition the tracking target of the vehicle balance parameter from the current balance parameter to the desired balance parameter.

[0079] Among them, the preset transition rate is positively correlated with the rate of change of road curvature of the vehicle's driving position in the prediction time domain.

[0080] S403, based on the vehicle balance parameters at each time point, constructs a sequence of target state variables.

[0081] It should be noted that the above-mentioned difference meets the preset condition, which means the difference can be greater than a preset difference threshold. The target state variable sequence constructed in this way can, in scenarios with small deviations and small differences, allow the tracking target of the vehicle balance parameters in all steps within the prediction time domain to directly take the desired balance parameters, enabling the model predictive controller to quickly converge to the steady-state target and improve control efficiency. Simultaneously, in scenarios with large deviations and large differences, the tracking target of the vehicle balance parameters generates a sequence of gradual changes according to a preset transition rate, ensuring a smooth transition of vehicle attitude. The transition rate is positively correlated with the road curvature, improving control efficiency while ensuring the vehicle does not rollover or become unstable.

[0082] In other words, the embodiments of this disclosure can directly track the desired balance parameter when the difference is small, avoiding invalid transitions, while adopting a gradual approximation strategy when the difference is large to prevent abrupt attitude changes, thus resolving the contradiction between smoothness and efficiency in the control process.

[0083] For example, other vehicle state parameters that may be involved in the target state quantity sequence, such as lateral displacement, heading, longitudinal velocity, and yaw rate, can be tracked relatively smoothly by directly combining the current vehicle state information with the trajectory decision information upstream of the autonomous vehicle. Therefore, no additional smoothing processing is required for these parameters, and this disclosure will not elaborate on them.

[0084] The above embodiments, in combination, provide a detailed description of the complete drift control process from feedforward control to feedback control disclosed herein. It is understood that the above control methods can be used to control a vehicle already in a drifting state, as well as to control a vehicle about to enter a drifting state at a target position (pre-aiming point).

[0085] Next, we will explain in detail the triggering conditions for a vehicle in normal driving mode to enter a drift driving state at a target location.

[0086] Specifically, for a vehicle currently in normal driving mode, if the vehicle meets the conditions for triggering drift mode, it can be determined that the vehicle has entered drift mode. At this time, the aforementioned S102 and S103 can be executed to control the vehicle to drift.

[0087] The drift mode trigger conditions include at least one of the following: The predicted wheel sideslip parameter of the vehicle at the target position is greater than the wheel sideslip parameter threshold. The wheel sideslip parameter threshold is determined based on the wet and slippery data at the target position, and the predicted wheel sideslip parameter is determined based on the vehicle's current balance parameters combined with the road information at the target position. The wheel sideslip parameter is used to characterize the trend of the vehicle's steering attitude change. The deviation between the vehicle's current attitude information and the target attitude information meets the preset deviation condition.

[0088] It should be noted that the triggering conditions for the two drift modes described above are set separately from two aspects: actively guiding vehicles with a tendency to slip and become unstable to drift (hereinafter referred to as "active drift") and guiding vehicles that have already passively slipped and become unstable to drift (hereinafter referred to as "passive drift").

[0089] For example, the triggering condition for active drift corresponds to the aforementioned "the predicted wheel slip parameter of the vehicle at the target position is greater than the wheel slip parameter threshold". Here, the wheel slip parameter can be the rear wheel slip angle of the vehicle. This triggering condition can proactively determine whether the vehicle is about to become unstable by comparing the predicted value of the rear wheel slip angle at the target position with its physical limit, thereby actively intervening and smoothly guiding the vehicle into a controllable drift state before instability occurs.

[0090] Specifically, predict wheel slip parameters (rear wheel slip angle). It can be calculated using the following formula: ; in, The sideslip angle is the angle of the centroid. This is the distance from the rear axle to the center of mass. The desired velocity is given by the desired yaw rate. Calculations show that The road curvature at the target location can be obtained from the road information at the target location.

[0091] Wheel slip parameter threshold (rear wheel slip angle) It can be calculated using the following formula: ; in, The road surface adhesion coefficient at the target location can be obtained from the wet / slip data of the target location. For real-time calculation of the rear axle vertical load, Based on the improved brush tire model The calculated lateral stiffness of the rear wheel.

[0092] Predicting wheel slip parameters (rear wheel slip angle) Greater than the wheel slip parameter threshold (rear wheel slip angle) Under these circumstances, determine if the vehicle meets the conditions for triggering drift mode.

[0093] For example, the passive triggering condition corresponds to the above-mentioned "the deviation between the vehicle's current attitude information and the target attitude information meets the preset deviation condition". Here, the attitude information can be yaw rate, lateral position information, and heading information, etc. This triggering condition can quickly identify and switch to drift mode for stabilization control when it is detected that the vehicle has begun to exhibit undesirable oversteer or abnormal attitude due to external interference (such as low-friction road surface), so as to prevent the situation from deteriorating.

[0094] Specifically, the vehicle's current attitude information includes: current yaw rate; the target attitude information includes: target yaw rate; and the preset deviation conditions include: the current yaw rate and the target yaw rate are in the same direction, and the current yaw rate is greater than a yaw rate threshold; and / or The vehicle's current attitude information includes: current lateral position information and current heading information. The target attitude information includes: target lateral position information and target heading information. The preset deviation conditions include: the lateral position error and the heading error are in opposite directions, and the heading error is greater than the heading error threshold. The lateral position error is determined based on the current lateral position information and the target lateral position information, and the heading error is determined based on the current heading information and the target heading information.

[0095] Among them, the yaw rate threshold is positively correlated with the target yaw rate.

[0096] In other words, the passive trigger condition can be expressed as: the current yaw rate. yaw rate of the target The directions are the same, and ,in A coefficient greater than 1; and / or, lateral error information. With heading error information The directions are opposite, and Greater than the heading error threshold .

[0097] When a vehicle meets any of the above passive triggering conditions, it can meet the drift mode triggering conditions.

[0098] This disclosure innovatively designs a dual triggering mechanism combining active and passive methods. The active triggering mechanism, based on a tire mechanics model, predicts instability before it occurs and smoothly intervenes before the vehicle's dynamic boundaries are reached, preventing problems before they arise. The passive triggering mechanism, based on real-time yaw and heading error monitoring, responds rapidly to correct course and stabilize the vehicle when it is about to become unstable due to sudden disturbances (such as driving over localized ice or potholes). This dual protection mechanism of "active prevention + passive emergency response" ensures that the system has a strong ability to cope with various anticipated and unexpected unstable conditions, greatly enhancing the system's robustness and driving safety.

[0099] In some embodiments, the wet slip data used to determine the wheel slip parameter threshold can be obtained by: generating preliminary wet slip information for the driving area based on historical wet slip data of the vehicle's driving area; correcting the preliminary wet slip information based on current environmental information to generate current wet slip information for the driving area; and determining wet slip data matching the target location from the current wet slip information. The environmental information includes at least one of weather information and road surface condition information.

[0100] For example, historical slippery data of the vehicle's driving area can be obtained to provide a basis for preliminary slippery information. For instance, data such as the road surface adhesion coefficient, wheel slip rate, and driving trajectory of each vehicle in the fleet when driving in this area can be combined with corresponding location coordinates, time, weather, and other related information to form a structured historical database.

[0101] Subsequently, weights can be assigned based on the similarity of vehicle operating conditions (such as vehicle speed and load) at the time of data collection. The closer the operating conditions are to the current requirements, the higher the weight, thus performing a weighted calculation on the historical slippery data of multiple vehicles. Then, a clustering algorithm is used to classify and integrate discrete data within the same area, eliminating single-vehicle estimation errors. Finally, based on the weighted clustering results, preliminary slippery information for the driving area is generated. This preliminary slippery information can be displayed in map form, thus forming a preliminary slippery data map covering the driving area.

[0102] For example, preliminary slippery information can be stored on a cloud server, and the current environmental information of the driving area can also be synchronized to the cloud server in real time. For instance, after the water truck finishes its work, information such as the truck's operating trajectory, watering time, and water volume can be sent to the cloud server, allowing the cloud server to correct the preliminary slippery information stored therein in real time based on the truck's operating status. As another example, real-time weather information (e.g., rainfall / snowfall information, air humidity, etc.) can be synchronized to the cloud server in real time, allowing the cloud server to correct the preliminary slippery information stored therein in real time based on the weather information.

[0103] Similarly, by fusing multi-source environmental information, the initial slippery information can be verified and corrected to generate current slippery information that fits the current operating conditions. This current slippery information can be represented as a slippery data map corresponding to the current driving area, ensuring that the data on the map is consistent with the real-time slippery road surface conditions.

[0104] For example, when a vehicle needs to obtain wet and slippery data of a target location, wet and slippery data matching the target location can be determined from the current wet and slippery information that is dynamically updated with environmental information, thereby completing the determination of the wheel slip parameter threshold of the target location.

[0105] This disclosed embodiment transcends the limitations of single-vehicle perception by innovatively incorporating cloud-based big data processing technology when acquiring slippery road data. The system integrates historically estimated slippery road data from multiple vehicles in a fleet, and verifies and corrects this data by combining it with real-time information such as water spraying and weather conditions, generating and dynamically updating a high-precision road slippery road data map. This allows the control system to accumulate and utilize historical operational experience, providing more accurate environmental perception for triggering drift control, thereby continuously optimizing control performance as operational data accumulates.

[0106] It is worth noting that in practical applications, step S101 will be executed regardless of whether the vehicle meets the aforementioned drift mode triggering conditions. This is to ensure the real-time nature of vehicle control command issuance. If the expected balance parameters and feedforward control quantity are calculated only after drifting is triggered, the vehicle's large inertia will cause a delay in the issuance of control commands (the calculation and execution require a control cycle). By this time, the vehicle may have already entered an unstable trend due to accumulated attitude deviations, and further adjustments could easily lead to rollover or trajectory deviation. Calculating the expected balance parameters and feedforward control quantity beforehand is equivalent to preparing a "control plan" for the drift mode before triggering the judgment. Once the triggering conditions are met, the feedforward control quantity can be directly issued as the initial control command without waiting for model calculation, significantly shortening the response time and adapting to the "transient working condition requirements" of sharp curves and low-friction roads in mines.

[0107] In some embodiments, for a vehicle in drift mode, if the vehicle no longer meets the drift mode triggering conditions or the vehicle exits a curve, it is determined that the vehicle has ended drift driving, and the vehicle's current balance parameters are smoothly transitioned to preset balance parameters in normal driving state, so that the vehicle smoothly transitions from drift driving to normal driving.

[0108] The specific details regarding the drift mode trigger conditions can be found in the description of the previous embodiments, and will not be repeated here. It is understood that in this embodiment, "the vehicle no longer meets the drift mode trigger conditions" means that the vehicle no longer meets any of the aforementioned drift trigger conditions; that is, it does not meet either the active drift trigger conditions or the passive drift trigger conditions. In this case, the vehicle can be controlled to end the drift.

[0109] For example, whether a vehicle has exited a curve can be determined based on road information at pre-aiming points. When the absolute value of the road curvature at multiple consecutive pre-aiming points decreases from large to small and is less than a preset curvature threshold, it can be determined that the vehicle has exited a curve. Of course, it is also possible to determine whether a vehicle has exited a curve by combining the vehicle's real-time location with a road network map; this disclosure does not limit this approach.

[0110] Understandably, when the vehicle ends its drift and gradually transitions to normal driving, its feedforward and feedback control parameters will change. At this point, for the feedforward control parameters, the parameters from the drift state can be transitioned to those from the normal driving state within a preset transition time. For the feedback control parameters, the tracking target for the vehicle balance parameter (center of gravity sideslip angle) in the target state sequence can be set to zero, and a rolling optimization solution can be obtained through a model predictive controller.

[0111] Similar to the drift state, in order to avoid the vehicle attitude changing abruptly after contacting the drift state, the tracking target setting strategy can be determined based on the difference between the current center of gravity sideslip angle and the expected center of gravity sideslip angle (which is set to zero after the drift ends). This achieves the effect of pursuing a smooth transition under large deviations and pursuing control efficiency under small deviations.

[0112] Specifically, if the difference Then let the tracking target at each time step in the target state sequence be... If the difference Then let the tracking target at each time step in the target state sequence be... With transition rate The transition from the current centroid sideslip angle to 0. The transition rate is... With the rate of change of road curvature Positive correlation and For different centroid sideslip angle difference thresholds, .

[0113] This disclosure provides a two-way exit condition, which avoids attitude instability caused by premature exit and prevents the impact on normal driving efficiency by exiting too late, adapting to complex scenarios such as unpaved curves and low-friction surfaces in mines. Simultaneously, during the transition from drift to normal state, a gradual attenuation strategy can be used to control balance parameters and control quantities, avoiding sudden attitude changes and component impacts during exit, achieving a seamless transition between drift driving and normal driving.

[0114] Based on the same inventive concept, this disclosure also provides a drift control device, as shown in the following embodiment. Since the principle of this drift control device embodiment in solving the problem is the same as that described above... Figure 1 The method embodiments shown are similar, therefore the implementation of this drift control device embodiment can be found in the above description. Figure 1 The implementation of the method embodiments shown will not be repeated here.

[0115] Figure 5 A schematic diagram of a drift control device according to an embodiment of this disclosure is shown. Figure 5As shown, the drift control device 500 includes: a prediction module 501, a generation module 502, and a control module 503.

[0116] The prediction module 501 is used to predict the desired balance parameters and feedforward control parameters of the vehicle at the target location based on the road information and vehicle state information at the target location.

[0117] The generation module 502 is used to generate feedback control parameters with the goal of controlling the current balance parameters of the vehicle to the desired balance parameters.

[0118] The control module 503 is used to control the vehicle to drift at the target position based on feedforward control parameters and feedback control parameters.

[0119] In some embodiments, the drift control device 500 further includes a guidance module (not shown) for determining that the vehicle has entered a drift mode when the vehicle meets the drift mode triggering conditions, and continuing to execute the step of generating feedback control parameters with the goal of controlling the vehicle's current balance parameters to the desired balance parameters.

[0120] Drift mode is triggered by at least one of the following conditions: The predicted wheel sideslip parameter of the vehicle at the target position is greater than the wheel sideslip parameter threshold. The wheel sideslip parameter threshold is determined based on the wet and slippery data at the target position, and the predicted wheel sideslip parameter is determined based on the vehicle's current balance parameters combined with the road information at the target position. The wheel sideslip parameter is used to characterize the trend of the vehicle's steering attitude change. The deviation between the vehicle's current attitude information and the target attitude information meets the preset deviation condition.

[0121] In some embodiments, the drift control device 500 further includes an acquisition module (not shown) for generating preliminary slippery information of the driving area based on historical slippery data of the vehicle driving area; correcting the preliminary slippery information based on current environmental information to generate current slippery information of the driving area; wherein the environmental information includes at least one of weather information and road surface condition information; and determining slippery data matching the target position from the current slippery information.

[0122] In some embodiments, the vehicle's current attitude information includes: current yaw rate; the target attitude information includes: target yaw rate; and the preset deviation conditions include: the current yaw rate and the target yaw rate are in the same direction, and the current yaw rate is greater than a yaw rate threshold; and / or The vehicle's current attitude information includes: current lateral position information and current heading information. The target attitude information includes: target lateral position information and target heading information. The preset deviation conditions include: the lateral position error and the heading error are in opposite directions, and the heading error is greater than the heading error threshold. The lateral position error is determined based on the current lateral position information and the target lateral position information, and the heading error is determined based on the current heading information and the target heading information.

[0123] In some embodiments, the control module 503 is used to superimpose the control quantity shown by the feedforward control parameters and the control quantity shown by the feedback control parameters to generate a target control quantity; and to convert the target control quantity into a control signal so as to control the vehicle to drift at the target position.

[0124] In some embodiments, the generation module 502 is used to construct a target state quantity sequence, wherein the target state quantity sequence is used to show the tracking target of the vehicle state parameters at each time in the prediction time domain, the vehicle state parameters include vehicle balance parameters, and the tracking target of the vehicle balance parameters is related to the desired balance parameters; the vehicle state parameters at the current time and the target state quantity sequence are input into a pre-constructed model prediction controller, so that the model prediction controller generates feedback control parameters for the current time of the vehicle with the objective of minimizing the difference between the vehicle state parameters at each time and the corresponding tracking target.

[0125] In some embodiments, the generation module 502 is used to determine the difference between the current balance parameter and the desired balance parameter of the vehicle; if the difference meets a preset condition, starting from the current balance parameter of the vehicle, the tracking target of the vehicle balance parameter at each moment in the prediction time domain is generated sequentially according to a preset transition rate, so as to smoothly transition the tracking target of the vehicle balance parameter from the current balance parameter to the desired balance parameter; wherein, the preset transition rate is positively correlated with the rate of change of road curvature of the vehicle's driving position in the prediction time domain; and a target state quantity sequence is constructed based on the tracking target of the vehicle balance parameter at each moment.

[0126] In some embodiments, the guidance module is configured to determine that the vehicle has ended drifting if the vehicle no longer meets the drift mode triggering conditions or the vehicle exits a curve, and to smoothly transition the vehicle's current balance parameters to preset balance parameters in normal driving state, so that the vehicle smoothly transitions from drifting to normal driving.

[0127] In some embodiments, the prediction module 501 is used to construct a system state model of the vehicle, which is associated with the vehicle's longitudinal velocity, sideslip angle, yaw rate, and feedforward control parameters; input road information and state information into the system state model, and, under the condition that the rate of change of longitudinal velocity, the rate of change of sideslip angle, and the rate of change of yaw rate in the system state model are constrained to zero, determine the desired balance parameters and feedforward control parameters of the vehicle at the target position through the system state model, wherein the desired balance parameters include the sideslip angle.

[0128] Based on the same inventive concept, this disclosure also provides an unmanned vehicle for performing the drift control method described above.

[0129] In some embodiments, this disclosure also provides a computer-readable storage medium, which may be a readable signal medium or a readable storage medium. A program product capable of implementing the methods described above is stored thereon. In some possible embodiments, various aspects of this disclosure may also be implemented as a program product including program code, which, when run on a terminal device, causes the terminal device to perform the steps described in the "Exemplary Methods" section of this specification according to various exemplary embodiments of this disclosure.

[0130] In some embodiments, this disclosure also provides a computer program product, including a computer program that, when executed by a controller, implements the steps of the drift control method described above.

[0131] These computer programs or instructions can be applied to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device, enabling the processor of the general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to function as an apparatus for implementing the corresponding functions in the above method embodiments.

[0132] Furthermore, the specific implementation form of the computer program product is not limited in the embodiments of this application. In some embodiments, the computer program product may be implemented as an application (APP), a mini-program, a PC client, a program module, a plug-in, an installation package, a software development kit (SDK), an image file of an optical disc (such as an ISO file), a plug-in, or software in the form of Software as a Service (SaaS), etc., but is not limited to these.

[0133] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the appended claims.

Claims

1. A drift control method, characterized in that, include: Based on road information and vehicle status information at the target location, predict the desired balance parameters and feedforward control parameters of the vehicle at the target location; With the goal of controlling the current balance parameters of the vehicle to the desired balance parameters, feedback control parameters are generated; Based on the feedforward control parameters and the feedback control parameters, the vehicle is controlled to drift at the target position.

2. The method according to claim 1, characterized in that, Before generating feedback control parameters with the goal of controlling the vehicle's current balance parameters to the desired balance parameters, the method further includes: If the vehicle meets the drift mode triggering conditions, it is determined that the vehicle has entered drift mode, and the step of generating feedback control parameters with the goal of controlling the current balance parameters of the vehicle to the desired balance parameters continues to be executed; The drift mode triggering condition includes at least one of the following: The predicted wheel sideslip parameter of the vehicle at the target position is greater than the wheel sideslip parameter threshold, wherein the wheel sideslip parameter threshold is determined based on the slippery data of the target position, and the predicted wheel sideslip parameter is determined based on the current balance parameter of the vehicle combined with the road information of the target position. The wheel sideslip parameter is used to characterize the trend of the vehicle's steering attitude change. The deviation between the vehicle's current attitude information and the target attitude information satisfies a preset deviation condition.

3. The method according to claim 2, characterized in that, The slippery data of the target location is obtained in the following manner: Based on historical slippery data of the vehicle's driving area, preliminary slippery information of the driving area is generated; Based on the current environmental information, the preliminary slippery information is corrected to generate the current slippery information of the driving area; wherein, the environmental information includes at least one of weather information and road surface condition information; Determine the slippery data that matches the target location from the current slippery information.

4. The method according to claim 2, characterized in that, The vehicle's current attitude information includes: current yaw rate; the target attitude information includes: target yaw rate; and the preset deviation conditions include: the current yaw rate and the target yaw rate are in the same direction, and the current yaw rate is greater than a yaw rate threshold; and / or The vehicle's current attitude information includes: current lateral position information and current heading information. The target attitude information includes: target lateral position information and target heading information. The preset deviation conditions include: the lateral position error and the heading error are in opposite directions, and the heading error is greater than the heading error threshold. The lateral position error is determined based on the current lateral position information and the target lateral position information, and the heading error is determined based on the current heading information and the target heading information.

5. The method according to claim 1, characterized in that, Based on the feedforward control parameters and the feedback control parameters, the vehicle is controlled to drift at the target position, including: The control quantity shown by the feedforward control parameters is superimposed with the control quantity shown by the feedback control parameters to generate the target control quantity; The target control quantity is converted into a control signal to control the vehicle to drift at the target position.

6. The method according to claim 1, characterized in that, The step of generating feedback control parameters with the goal of controlling the current balance parameters of the vehicle to the desired balance parameters includes: Construct a target state quantity sequence, wherein the target state quantity sequence is used to show the tracking target of the vehicle state parameters at each time in the prediction time domain, the vehicle state parameters include vehicle balance parameters, and the tracking target of the vehicle balance parameters is related to the expected balance parameters; The vehicle state parameters at the current moment and the target state sequence are input into a pre-built model prediction controller, so that the model prediction controller generates the feedback control parameters at the current moment of the vehicle with the objective of minimizing the difference between the vehicle state parameters at each moment of the vehicle and the corresponding tracking target.

7. The method according to claim 6, characterized in that, Construct the target state variable sequence, including: Determine the difference between the current balance parameters of the vehicle and the desired balance parameters; When the difference meets the preset conditions, starting from the current balance parameter of the vehicle, the tracking target of the vehicle balance parameter at each moment in the prediction time domain is generated by sequentially increasing the preset transition rate, so as to smoothly transition the tracking target of the vehicle balance parameter from the current balance parameter to the desired balance parameter; wherein, the preset transition rate is positively correlated with the rate of change of road curvature of the vehicle's driving position in the prediction time domain. Based on the vehicle balance parameters at each time point, a target state quantity sequence is constructed.

8. The method according to claim 2, characterized in that, The method further includes: If the vehicle no longer meets the drift mode triggering conditions or the vehicle exits the curve, the vehicle is determined to have ended drifting, and the vehicle's current balance parameters are smoothly transitioned to the preset balance parameters in normal driving state, so that the vehicle smoothly transitions from drifting to normal driving.

9. The method according to claim 1, characterized in that, The prediction of the vehicle's desired balance parameters and feedforward control parameters at the target location, based on road information and vehicle state information, includes: A system state model of the vehicle is constructed, which is associated with the vehicle's longitudinal velocity, center of gravity sideslip angle, yaw rate, and feedforward control parameters. The road information and the state information are input into the system state model. Under the condition that the rate of change of longitudinal velocity, the rate of change of centroid sideslip angle and the rate of change of yaw rate in the system state model are constrained to zero, the desired balance parameters and feedforward control parameters of the vehicle at the target position are determined by the system state model. The desired balance parameters include the centroid sideslip angle.

10. An unmanned vehicle, characterized in that, Used to perform the drift control method as described in any one of claims 1 to 9.

Images